LLMs can automate the extraction of user stories from high‑fidelity mockups.
Inclusion of a Language Extended Lexicon (LEL) glossary in prompts boosts accuracy.
A case study compares model performance with and without the LEL glossary.
Results support the integration of AI into requirements engineering processes.
The approach may improve communication between end‑users and developers.
📖 Full Retelling
A team of researchers—including Diego Firmenich, Leandro Antonelli, Bruno Pazos, Fabricio Lozada, and Leonardo Morales—presented a study on how large language models (LLMs) can automatically generate user stories from high‑fidelity mockups. The paper was first submitted to arXiv on 19 Feb 2026 and later presented at the 28th Workshop on Requirements Engineering (WER 2025), illustrating how adding a glossary from the Language Extended Lexicon (LEL) to prompts significantly improves the accuracy and suitability of the extracted user stories, thereby advancing AI‑driven requirements engineering.
🏷️ Themes
Artificial Intelligence in Software Engineering, Requirements Engineering, Natural Language Processing, Agile Methods and Automation, Human‑Computer Interaction with Mockups
Entity Intersection Graph
No entity connections available yet for this article.
Deep Analysis
Why It Matters
Automating user story generation from mockups speeds up requirements engineering and improves communication between users and developers. The study shows that adding a domain glossary boosts accuracy, making AI a practical aid in agile workflows.
Context & Background
User stories are a core agile artifact used to capture functional requirements
High‑fidelity mockups enable end‑users to participate early in design
Large language models can interpret natural language prompts and visual inputs
What Happens Next
The technique is expected to be incorporated into requirement‑engineering tools and IDE plugins, encouraging broader adoption in industry. Future research will extend the approach to other artifacts such as use‑case diagrams and test cases.
Frequently Asked Questions
What is a user story?
A short, user‑centric statement describing a feature or need.
How does the glossary improve results?
It supplies domain‑specific terms that help the model understand context and generate more accurate stories.
Will this replace manual requirement writing?
No, it assists developers but still requires human review to ensure correctness and completeness.
Original Source
--> Computer Science > Software Engineering arXiv:2602.16997 [Submitted on 19 Feb 2026] Title: Exploring LLMs for User Story Extraction from Mockups Authors: Diego Firmenich , Leandro Antonelli , Bruno Pazos , Fabricio Lozada , Leonardo Morales View a PDF of the paper titled Exploring LLMs for User Story Extraction from Mockups, by Diego Firmenich and 4 other authors View PDF HTML Abstract: User stories are one of the most widely used artifacts in the software industry to define functional requirements. In parallel, the use of high-fidelity mockups facilitates end-user participation in defining their needs. In this work, we explore how combining these techniques with large language models enables agile and automated generation of user stories from mockups. To this end, we present a case study that analyzes the ability of LLMs to extract user stories from high-fidelity mockups, both with and without the inclusion of a glossary of the Language Extended Lexicon in the prompts. Our results demonstrate that incorporating the LEL significantly enhances the accuracy and suitability of the generated user stories. This approach represents a step forward in the integration of AI into requirements engineering, with the potential to improve communication between users and developers. Comments: 14 pages, 6 figures. Preprint of the paper published in the 28th Workshop on Requirements Engineering (WER 2025) Subjects: Software Engineering (cs.SE) ; Artificial Intelligence (cs.AI); Computation and Language (cs.CL) ACM classes: D.2.1; I.2.7; D.2.2 Cite as: arXiv:2602.16997 [cs.SE] (or arXiv:2602.16997v1 [cs.SE] for this version) https://doi.org/10.48550/arXiv.2602.16997 Focus to learn more arXiv-issued DOI via DataCite (pending registration) Journal reference: Proceedings of the 28th Workshop on Requirements Engineering (WER2025) Related DOI : https://doi.org/10.29327/1588952.28-10 Focus to learn more DOI linking to related resources Submission history From: Diego Firmenich [ view em...
